Introduction

This document is a fake epidemiological report using data provided by the Applied Epi group using a fake COVID linelist (data available on Github).

I have used these data to create various descriptive analysis. Most of the work here is inspired from the great Epidemiologist R handbook, but also from other resources that I have used to learn R. I use pipes and all of the tidyverse has I have learned in: R for Data Science and other resources available on Rstudio website.

For data visualisation I have decided to focuss on interactive graphs using the highcharter package a wrapper for the highchart library. I have been working with highcharter for some time now but I used it as an opportunity to learn more, and apply it to a real life scenarios. I do really enjoy the adds-on of interactivity in data analysis.

For Epidemiological analysis/description I have used what I knew from the various packages regarding epi: linelist, epikit, incidence, sitrep …

For spatial analysis, I have chosen to get out of ggplot, or tmap which I have used a lot, to focuss on leaflet as it was once more an occasion to use interactivity in real life scenario for data visualisation and I think this is very powerful.

This report is more like a aggregate of data analysis and visualisation, but I have tried to stick to an epidemiologically sound structure of PERSON, TIME & PLACE.

Overview

Between the 2020-01-27 and the 2021-07-19 (a period 540 days) a total of 81979 cases have been identified, of which 82101 (100.0%) are confirmed cases.

  • In Fulton County, and over this period, the attack rate is of 789.06 (CI 783.89–794.26) per 10 000.

  • Amongst all cases of known hospitalisation status, at least 5297 (10.7%) have been hospitalised, but 32482 (39.6%) still have an unknown status

  • Using only cases with known outcome, the Case Fatality Ratio is of 3.37% (CI 3.20–3.55), but importantly, 123 (0.1%) cases are under review, and 42302 (51.5%) of cases missing outcome status.

Person

Since the start of the outbreak there was, 43299 (52.7%) female cases and 38393 (46.8%) male cases.

The Age of cases ranges from 0 to 106, with Age group 20 - 30 the most represented, with 18609 (22.67%) cases.

Missing data

There is 409 (0.5%) cases with missing data on Sex, and 49 (0.1%) cases with missing data on Age.

Age and Sex distributions

Distribution of cases per Age groups and case gender
Age group male cases (n) % female cases (n) % unknown cases (n) %
< 10 1423 3.71 1398 3.23 18 4.40
10 - 20 4473 11.65 4679 10.81 49 11.98
20 - 30 8280 21.57 10243 23.66 86 21.03
30 - 40 7333 19.10 8200 18.94 82 20.05
40 - 50 5949 15.50 6551 15.13 50 12.22
50 - 60 5339 13.91 5587 12.90 45 11.00
60 - 70 3126 8.14 3274 7.56 28 6.85
70 + 2455 6.39 3345 7.73 39 9.54
unknown 15 0.04 22 0.05 12 2.93

Age-Sex pyramid

Hospitalisations

Overall Hospitalisation rate is of 10.68% (CI 10.41–10.95), but 32482 (39.6%) still have an unknown status.

Age group

The age group, 70 + had the greatest hospitalisation rate of 44.61% (CI 43.00–46.23).

Hospitalisation rates per Age groups
Age groups hospitalisations population Hosp. rate 95%-CI
< 10 61 2005 3.04 (2.38-3.89)
10 - 20 80 5677 1.41 (1.13-1.75)
20 - 30 372 10792 3.45 (3.12-3.81)
30 - 40 575 9389 6.12 (5.66-6.63)
40 - 50 667 7603 8.77 (8.16-9.43)
50 - 60 982 6588 14.91 (14.07-15.79)
60 - 70 936 3917 23.90 (22.59-25.26)
70 + 1622 3636 44.61 (43-46.23)
unknown 2 12 16.67 (4.7-44.8)

Sex

Hopsitalisations rates remained similar between Male ( 11.22% (CI 10.82–11.64) ) and Female ( 11.22% (CI 10.82–11.64) ).

Hospitalisation rates per Sex
Sex hospitalisations population Hosp. rate 95%-CI
male 2582 23003 11.22 (10.82-11.64)
female 2710 26524 10.22 (9.86-10.59)
unknown 5 92 5.43 (2.34-12.1)

COVID Deaths

Using only cases with known outcome, the Overall CFR is of 3.37% (CI 3.20–3.55).

Importantly, 123 (0.1%) cases are under review, while 42302 (51.5%) have unknown outcome.

Age group

Distribution of COVID CFR per Age group
Age groups deaths population CFR 95%-CI
< 10 0 1620 0.00 (0-0.24)
10 - 20 1 4630 0.02 (0-0.12)
20 - 30 6 8700 0.07 (0.03-0.15)
30 - 40 27 7517 0.36 (0.25-0.52)
40 - 50 43 6086 0.71 (0.52-0.95)
50 - 60 125 5148 2.43 (2.04-2.89)
60 - 70 252 3097 8.14 (7.23-9.15)
70 + 884 2871 30.79 (29.13-32.5)
unknown 0 7 0.00 (0-35.43)

< 10 was the age group with the lowest CFR: 0.00% (CI 0.00–0.24). While 70 + was the age group with the greatest CFR: 30.79% (CI 29.13–32.50)

Sex

Distribution of COVID CFR per Sex
Sex deaths population CFR 95%-CI
male 705 18440 3.82 (3.56–4.11)
female 633 21183 2.99 (2.77–3.23)
unknown 0 53 0.00 (0.00–6.76)

Symptoms distribution

Amongst all cases, symptoms were distributed as follow:

Distribution of symptoms
value n %
Fever
yes 15127 18.42
no 33951 41.35
unknown 33023 40.22
Subjective Fever
yes 12712 15.48
no 30457 37.10
unknown 38932 47.42
Myalgia
no 29210 35.58
yes 19534 23.79
unknown 33357 40.63
Loss of Taste & Smell
unknown 51258 62.43
yes 12734 15.51
no 18109 22.06
Sore throat
yes 12516 15.24
no 36106 43.98
unknown 33479 40.78
Cough
yes 21943 26.73
unknown 32684 39.81
no 27474 33.46
Headache
yes 21675 26.40
no 27196 33.13
unknown 33230 40.47

And this was related to one another as follow:

Time

Main Epicurve

The peak of the outbreak was reached on week 2020-W53, with a total of 3615 cases. 122 cases had missing dates, and they were excluded from the epicurves.

Epicurves by groups

Weekly Attack Rates, Hospitalisation rates and CFR

Place

Cases map

Cases have been detected throughout Fulton county, especially in major cities.

Chloropleths maps

Peak of Outbreak

Below are mapped the smaller regions of Fulton county (<1000 habitants) which AR are really important (mean: 149.83) due to small size populations, mean: 306.37. This was done to allow a more nuanced mapping of other bigger regions by removing outliers.

At the peak of the outbreak: 2020-W53, the region with the greatest Attack rate was 3.0024^{4} with an Attack rate of 167.2240803 (71.63–385.42) per 10 000

Months and week

Maps of big regions of Fulton county by Week of Outbreak.